DESCRIPTION: This project will examine the effects of post-marital unions (remarriage and cohabitation) on the well-being of children. Available evidence suggests that children find the adjustment to step-family life difficult, yet most studies do not carefully distinguish between the lingering effects of divorce and emergent effects of remarriage. Furthermore, although most divorced adults enter new relationships, many of the new unions are non- marital. Yet, very little is known about how these often transitory relationships affect child well-being. In the proposed study the investigators plan to address the following key questions: 1) does remarriage improve or worsen child well-being in the aftermath of divorce compared with living with an opposite sex partner or remaining in a single parent family, net of selectivity factors?; 2) are effects of post-marital unions similar for younger versus older children? girls versus boys?; 3) what are the moderating influences by which post-marital unions affect child well-being?; and 4) do difficulties associated with remarriage and cohabitation diminish with time? Data from the National Longitudinal Survey of Youth - Child Supplement (NLSY- CS) (i.e., 1986, 1988, 1990, 1992, and 1994 waves) will be pooled to examine the effects of post-marital unions on child well-being. The sample will be limited to children ranging in age from five to 16 in 1994. Longitudinal information from follow-up assessments will be used to explore children s adjustment over time and the factors that enhance or undermine it. Dependent variables to be examined include the Behavior Problems Index; the mathematics and reading sub-tests of the Peabody Individual Achievement Tests; and for children at least ten years old at the time of the assessment, Harter s (1982) measure of children s self perception and academic self-worth, and children s propensity to act out and break rules, such as staying out late without permission and skipping school. The statistical analyses will begin with generalized least squares regression models that take account of maternal self-selection into marital disruption. Next, because some of the potentially important sources of diversity in post-marital unions logically exist only for children whose parents are in a union, such as the presence of step- siblings, models will be estimated only for those children whose mothers are remarried or cohabiting. Bivariate probit specification will be used to address this two-tiered sample selection problem.